Skeletal muscle AMP-activated protein kinase γ1^H151R overexpression enhances whole body energy homeostasis and insulin sensitivity

Abstract

AMP-activated protein kinase (AMPK) is a major sensor of energy homeostasis and stimulates ATP-generating processes such as lipid oxidation and glycolysis in peripheral tissues. The heterotrimeric enzyme consists of a catalytic α-subunit, a β-subunit that is important for enzyme activity, and a noncatalytic γ-subunit that binds AMP and activates the AMPK complex. We generated a skeletal muscle Cre-inducible transgenic mouse model expressing a mutant γ₁-subunit (AMPKγ1^H151R), resulting in chronic AMPK activation. The expression of the predominant AMPKγ3 isoform in skeletal muscle was reduced in extensor digitorum longus (EDL) muscle (81–83%) of AMPKγ1^H151R transgenic mice, whereas the abundance and phosphorylation of the AMPK target acetyl-CoA carboxylase was increased in tibialis anterior muscle. Glycogen content was increased 10-fold in gastrocnemius muscle. Whole body carbohydrate oxidation was increased by 11%, and whereas glucose tolerance was unaffected, insulin sensitivity was increased in AMPKγ1^H151R transgenic mice. Furthermore, perigonadal white adipose tissue mass and serum leptin were reduced in female AMPKγ1^H151R transgenic mice by 38 and 51% respectively. Conversely, in male AMPKγ1^H151R transgenic mice, food intake was increased (14%), but body weight and body composition were unaltered, presumably because of increased energy expenditure. In conclusion, transgenic activation of skeletal muscle AMPKγ1 in this model plays an important sex-specific role in skeletal muscle metabolism and whole body energy homeostasis.

the 5′amp-activated protein kinase (AMPK) is the main sensor of cellular energy status and thereby plays a central role in the regulation of metabolism. The role of AMPK in exercise-induced metabolic adaptations and a variety of signaling pathways enhancing glucose and lipid metabolism has raised interest in this protein kinase as a drug target for the treatment of type 2 diabetes and obesity (33, 44). AMPK stimulates ATP-generating processes such as glucose uptake, glycolysis, and fatty acid oxidation and inhibits ATP-consuming pathways such as fatty acid or triglyceride synthesis to restore energy balance (9, 20). AMPK is a heterotrimeric enzyme consisting of one catalytic (α₁ or α₂) and two regulatory subunits (β₁ or β₂ and γ₁, γ₂, or γ₃) (15, 18). Binding of AMP, and most likely ADP, to the γ-subunit induces an allosteric activation of the AMPK complex and protection of the Thr¹⁷² residue on the α-subunit from dephosphorylation, therefore maintaining the enzyme in an active state (12, 19, 38). AMPK activation promotes GLUT4 translocation to the plasma membrane (25) and increases cellular glucose uptake in an insulin-independent manner even in severely insulin-resistant type 2 diabetic patients (24, 32). These preclinical observations point toward AMPK as therapeutic target to enhance insulin sensitivity.

Downregulation of AMPK activity in peripheral tissues reduces exercise capacity, worsens glucose tolerance, and promotes obesity and diabetes in several AMPK isoform-knockout or kinase-dead mouse models (7, 16, 39, 41, 42). Conversely, the presence of a naturally occurring activating mutation in the γ₃-subunit of Hampshire pigs (R220Q) and humans (R225W) promotes glycogen accumulation and increases citrate synthase activity as well as oxidative capacity in skeletal muscle (3, 13, 30). Mutant forms of the γ₂-subunit in mouse models and humans increase glycogen storage in cardiomyocytes and cause Wolff-Parkinson-White syndrome (4, 10). Several transgenic mouse models harboring AMPK-activating point mutations support a role for the γ-subunits in the regulation of glucose and lipid metabolism. The three AMPKγ subunits share four conserved cystathionine β-synthase (CBS) domains but differ in their NH₂-terminal sequence (1, 12). Two corresponding murine mutations, AMPKγ1^R70Q and AMPKγ3^R225Q, are both located in the CBS1 domain (7, 8) (Fig. 1A). AMPKγ3^R225Q transgenic mice have increased skeletal muscle glycogen content and are protected against high-fat diet-induced triglyceride accumulation and insulin resistance (7, 17, 27, 43). Despite evidence that AMPKγ3 is the major γ-subunit expressed in skeletal muscle, transgenic mice expressing an AMPKγ1 gain-of-function mutation (AMPKγ1^R70Q) also have increased skeletal muscle glycogen storage (8), indicating that AMPK activation by modifying either the γ₁- or the γ₃-subunit may protect against the development of insulin resistance in obesity or type 2 diabetes. However, the impact of an activating mutation in the AMPKγ1 subunit in skeletal muscle and effects on whole body glucose homeostasis and energy homeostasis in female and male mice has not been investigated. This is of particular interest given that the sensitivity of AMPK to activating compounds can depend on the heterotrimer composition and the presence of the γ₁-subunit (22).

Fig. 1.

AMP-activated protein kinase (AMPK) mutations and gene expression in skeletal muscle. A: a schematic representation of the AMPKγ1^R70Q and AMPKγ3^R225Q mutations in the cystathionine β-synthase (CBS)1 domain and the AMPKγ1^H151R mutation in the CBS2 domain. mRNA expression of the AMPKγ1^H151R transgene, endogenous AMPKγ1 (Prkag1), and AMPKγ3 (Prkag3) was quantified in extensor digitorum longus (EDL) and soleus (SOL) muscle of 22-wk-old wild-type (WT; open bars) and AMPKγ1^H151R transgenic mice (black bars). B and D: the quantification of AMPKγ1^H151R transgene expression in EDL and soleus from male (B) and female mice (D) is shown in threshold cycle (C_T) values since the transgene is not detectable in WT mice. C and E: endogenous Prkag1 and Prkag3 expression in EDL and soleus from male (C) and female mice (E) is shown as fold change to expression in WT mice. Results are means ± SE. *P < 0.05 and ***P < 0.001 vs. respective WT mice; n = 5–11 mice. ND, not detectable.

Activating mutations of the AMPK complex generally bypass the requirement for AMP to bind the γ-subunit and activate the enzyme complex. In yeast, the absence of His¹⁵¹ in the CBS2 domain of the structural AMPKγ1 yeast homolog snf4p renders the complex insensitive to AMP (1). Additionally, the introduction of the murine homolog AMPKγ1^H150R via adenoviral transfection into the hypothalamus of mice promoted a strong activation of AMPK independent of the intracellular ATP/AMP ratio (31). Thus, this mutation is considered an activating mutation in vitro and in vivo. Collectively, these studies suggest that the His¹⁵¹ residue of the γ₁-subunit may play a role in the regulation of glucose homeostasis. In this study, we investigated the impact of AMPKγ1 activation in skeletal muscle on whole body metabolism as well as gene expression in other peripheral tissues, including adipose tissue. We generated a transgenic mouse model carrying an activating mutation at CBS2 His¹⁵¹ of the human homolog of the AMPKγ1 subunit in skeletal muscle. We provide evidence that skeletal muscle AMPKγ1 activation improves whole body insulin sensitivity and energy homeostasis in a sex-specific manner.

MATERIALS AND METHODS

Animals.

Mice expressing Cre in myosin light chain (MLC)1 in skeletal muscle (11) (kindly provided by Steven Burden, New York University Medical Center, New York, NY) were mated with mice carrying the Cre-inducible transgene AMPKγ1^H151R. The transgene is under control of the β-actin promoter (2) and contains the mutated human PRKAG1 gene sequence located downstream of a stop sequence flanked by two loxP sites. This allows the expression of the transgene in the presence of the Cre recombinase. Histidine 151 in the human protein sequence refers to the previously described murine-activating mutation of H150 in the CBS2 domain of the AMP-binding site of AMPKγ1 (31). We studied female and male floxed (fl/fl) AMPKγ1^H151R MLC1-Cre (AMPKγ1^H151R) mice and wild-type littermates bred on a mixed C57BL/6J background. Animals were maintained at 24°C and on a 12:12-h light-dark cycle with free access to standard chow and water. All animal protocols were approved by the regional animal ethics committee at Stockholm North.

Metabolic cages.

Mice (16–22 wk of age) were acclimated for 24 h in single cages. Subsequently, mice were monitored for 2 days in the Oxymax Lab Animal Monitoring System (Comprehensive Laboratory Animal Monitoring System; Columbus Instruments, Columbus, OH) with ad libitum access to food and one night without access to food (12-h fasting). Food intake, drinking volume, O₂ consumption (V̇o₂), CO₂ production (V̇co₂), respiratory exchange ratio (RER), heat, and movement were measured continuously.

Glucose tolerance test.

Mice (17 wk of age) were fasted for 4 h, and thereafter, 2 g/kg glucose was administered intraperitoneally. Plasma glucose was measured in whole blood at 0, 15, 30, 60, and 120 min (One Touch Ultra 2 glucose meter; Lifescan, Milpitas, CA). Serum insulin concentration was measured at 0 and 15 min using an ELISA with mouse insulin as a standard (Crystal Chem, Chicago, IL).

Hyperinsulinemic euglycemic clamp.

Jugular vein catheterization was performed under isoflurane anesthesia with carprofen analgesic treatment. Animals were left to recover for at least 4 days in single cages. Using a constant jugular vein infusion of [3-³H]glucose (2.5 μCi bolus and a flow rate of 0.04 μCi/min), glucose turnover rate was measured in the basal state and under hyperinsulinemic euglycemic clamp conditions in 4-h-fasted conscious mice. Basal glucose utilization and hepatic glucose production was assessed 50–70 min after the start of the tracer infusion right before the insulin infusion was started. The clamp was started with a priming dose of insulin (17.5 mU/kg, Actrapid; Novo Nordisk, Bagsvaerd, Denmark), followed by a constant infusion of insulin at a rate of 1.75 mU·kg⁻¹·min⁻¹. At steady state (∼85 min after the start of the insulin infusion), blood samples were collected and whole body glucose utilization was measured. Hepatic glucose production was calculated by subtracting the glucose infusion rate from the glucose utilization. Additional blood samples were taken at the basal and clamp states to determine serum insulin concentrations by ELISA (Crystal Chem). Animals were euthanized by an overdose of pentobarbital sodium.

Body composition.

Total lean and fat mass was assessed in conscious mice (19 wk of age) using the EchoMRI-100 system (Echo Medical Systems, Houston, TX).

Tissue collection.

Mice (22 wk of age) were fasted for 4 h and anesthetized via an intraperitoneal injection of 16 μl/g body wt 2.5% avertin (2,2,2-tribromo ethanol and tertiary amyl alcohol). Hindlimb muscles, perigonadal white adipose tissue (WAT), heart, and brown adipose tissue (BAT) were snap-frozen in liquid nitrogen and stored at −80°C. Blood was collected, and serum was stored at −80°C.

Glycogen and triglyceride content.

Glycogen content in gastrocnemius muscle and heart was measured in 4-h-fasted mice (22 wk of age) using a glycogen assay kit according to the manufacturer's protocol (Abcam, Cambridge, UK). Triglyceride was extracted from tibialis anterior muscle using a heptane-isopropanol (3:2) mix, and triglyceride concentration was determined using the Trig/GB kit (Roche Diagnostics, Indianapolis, IN).

Leptin measurement.

Serum leptin concentration was assessed by an ELISA (Crystal Chem) in blood samples collected on the day of terminal dissection from 4-h-fasted mice (22 wk of age).

Tissue preparation for protein analysis.

Tibialis anterior muscle samples were homogenized in ice-cold homogenization buffer (137 mM NaCl, 2.7 mM KCl, 1 mM MgCl₂, 5 mM Na₄P₂O₇, 10 mM NaF, 1% Triton X-100, 10% glycerol, 20 mM Tris, pH 7.8, 1 mM EDTA, 0.2 mM PMSF, 0.5 mM Na₃VO₄, and 1× protease inhibitor cocktail; Merck Millipore, Nottingham, UK) using a TissueLyzer (Qiagen, Hilden, Germany). The samples were rotated for 1 h at 4°C and subjected to centrifugation (15 min at 12,000 g at 4°C) before the supernatant was collected. Protein content in the supernatant was measured using the Pierce BCA Protein assay kit (Thermo Scientific, Waltham, MA). Lysates were adjusted to equal protein concentrations and heated in Laemmli buffer for 20 min at 56°C.

Western blot analysis.

Proteins were separated by SDS-PAGE using Criterion XT Precast gels (Bio-Rad, Hercules, CA) and transferred to polyvinylidene difluoride membrane (Immobilon-P; Millipore, Billerica, MA). Membranes were blocked for 1 h at room temperature in TBST buffer (10 mmol/l Tris-base, 150 mmol/l NaCl, 0.02% Tween) containing 5% milk. After overnight incubation with the primary antibody, membranes were incubated with a horseradish peroxidase-conjugated secondary antibody (goat anti-rabbit or rabbit anti-mouse; Bio-Rad) in TBST buffer with 1% milk for 1 h at room temperature, and bands were visualized using enhanced chemiluminescence (GE Healthcare, Little Chalfont, UK). Bands were quantified using Quantity One 1-D analysis software (Bio-Rad) and normalized to total protein staining with Ponceau S (Sigma Aldrich, St. Louis, MO). The following antibodies from Cell Signaling Technology (Danvers, MA) were used: phospho-acetyl-CoA carboxylase (ACC)-α/β (Ser⁷⁹), no. 3661; ACC-α/β, no. 3662; phospho-AMPKα1/2 (Thr¹⁷²), no. 2531; AMPKα1/2, no. 2532; AMPKγ1, no. 4187; glycogen synthase, no. 3893; phospho-glycogen synthase (Ser⁶⁴¹), no. 3891. The GLUT4 antibody (no. 07-1404) was from Merck Millipore (Nottingham, UK). The AMPKγ3 antibody was kindly provided by Dr. Grahame Hardie (University of Dundee, Dundee, UK), and the hexokinase 2 antibody was kindly provided by Dr. Oluf Pedersen (University of Copenhagen, Copenhagen, Denmark). The GLUT1 antibody was a kind gift from Dr. Geoffrey Holman (University of Bath, Bath, UK).

RNA extraction and gene expression analysis.

Gene expression was measured in extensor digitorum longus (EDL) and soleus muscle and perigonadal WAT and BAT from 4-h-fasted mice. RNA was extracted using TRIzol reagent (Life Technologies, Carlsbad, CA) and chloroform and transcribed into cDNA using the High Capacity cDNA Reverse Transcription Kit (Life Technologies). TaqMan assays for Ppia (Mm02342430_g1), PRKAG1 (Hs00176952_m1), Prkag1 (Mm00450298_g1), Prkag3 (Mm00463997_m1), Ucp-1 (Mm01244861_m1), and Pdk4 (Mm01166879_m1) were purchased from Life Technologies. SYBR green primer sets for β₃-adrenergic receptor (Adrb3; forward 5′-GGC CCT CTC TAG TTC CCA G-3′; reverse 5′TAG CCA TCA AAC CTG TTG3′), Actb (forward 5′-ACA CCC CAG CCA TGT ACG TAG-3′; reverse 5′AGG CAT ACA GGG ACA GCA CAG3′), and Tbp [forward 5′-GAA GCT GCG GTA CAA TTC CAG-3′; reverse 5′-CCC CTT GTA CCC TTC ACC AAT-3′ (described in Ref. 34)] were purchased from Sigma-Aldrich. Relative changes in gene expression were analyzed using the 2^−ΔΔC_T method.

Statistics.

Results are presented as means ± SE. Differences were determined using Student's t-test or two-way ANOVA where applicable. For post hoc analysis, Bonferroni's test was used. Differences were considered statistically significant at P < 0.05.

RESULTS

AMPKγ1^H151R alters endogenous AMPK subunit expression.

To confirm the expression of the skeletal muscle promoter-driven AMPKγ1-activating transgene, we quantified mRNA in tissues from male and female wild-type and transgenic mice. The AMPKγ1^H151R transgene was detectable in EDL and to a lesser extent (represented by higher C_T values) in soleus muscle from male and female AMPKγ1^H151R mice but not wild-type mice (Fig. 1, B and D). Overexpression of AMPKγ1^H151R significantly reduced the expression of endogenous AMPKγ1 (Prkag1) in EDL (30% in males and 43% in females) but not soleus muscle (Fig. 1, C and E). Expression of AMPKγ3 (Prkag3), the main AMPKγ isoform in glycolytic muscle, was reduced in EDL muscle (81% in males and 83% in females) and to a lesser extent in soleus muscle of male (48%; Fig. 1C) but not female transgenic mice (Fig. 1E). The increased AMPKγ1 and reduced AMPKγ3 expression was also noted at the protein level in tibialis anterior muscle (>4.6-fold increase in AMPKγ1 and 55% reduction of AMPKγ3 in male and 71% reduction in female mice; Fig. 2, A and B). AMPKα Thr¹⁷² phosphorylation was increased (69%) in female but not male AMPKγ1^H151R transgenic mice. Nevertheless, total protein abundance of AMPKα (20% in males and 31% in females), as well as ACC-α/β (86% in males and 96% in females), was increased in both female and male AMPKγ1^H151R transgenic mice (Fig. 2, A and B). Moreover, abundance of pACC Ser⁷⁹ (2.7-fold in males and 3.7-fold in females; Fig. 2, A and B), as well as the p-ACC/ACC ratio (50% in males and 92% in females; data not shown) was increased in both sexes, indicating that overexpression of AMPKγ1^H151R in skeletal muscle increases AMPK signal transduction.

Fig. 2.

Protein levels of AMPK and AMPK targets. Levels of total AMPKγ1, total AMPKγ3 (arrow), phosphorylated AMPKα1/2 Thr¹⁷², total AMPKα1/2, phosphorylated ACCα/β Ser⁷⁹, and total ACCα/β in tibialis anterior muscle from 22-wk-old male (A) and female mice (B) are shown in representative Western blots (left lanes: WT; right lanes: AMPKγ1^H151R). Quantification is presented in arbitrary units (AU) for WT (open bars) and AMPKγ1^H151R transgenic mice (black bars). Results are means ± SE. *P < 0.05, **P < 0.01, and ***P < 0.001; n = 4–8 mice.

Increased skeletal muscle glycogen content is not caused by chronic glycogen synthase activation.

Skeletal muscle glycogen content was increased more than 10-fold in gastrocnemius muscle from AMPKγ1^H151R transgenic mice (Fig. 3A). Despite this, glycogen synthase protein abundance in skeletal muscle was unaltered in male but slightly reduced in female AMPKγ1^H151R transgenic mice (11%; Fig. 3, B and C). Phosphorylation of glycogen synthase at Ser⁶⁴¹ was increased (2.5-fold in male and 2.6-fold in female mice), suggesting that chronic AMPKγ1^H151R overexpression reduces glycogen synthase activity. Glycogen levels in cardiac muscle were unaltered between AMPKγ1^H151R transgenic and wild-type mice (data not shown). Hexokinase 2 protein abundance was not increased in tibialis anterior muscle from male AMPKγ1^H151R transgenic mice (16%, P = 0.08). GLUT1 protein abundance was reduced in male (21%, Fig. 4A) but not in female (Fig. 4D) AMPKγ1^H151R transgenic mice. GLUT4 protein abundance was increased (34% in male and 37% in female mice; Fig. 4, B and E). Gene expression of Pdk4 was reduced in AMPKγ1^H151R transgenic mice (47% in male and 63% in female mice; Fig. 4, C and F).

Fig. 3.

Skeletal muscle glycogen content and regulation of glycogen synthase. A: intramuscular glycogen storage was assessed in gastrocnemius muscle from male and female WT (open bars) and AMPKγ1^H151R transgenic mice (black bars). B and C: protein level of total and phosphorylated (Ser⁶⁴¹) glycogen synthase (GS) is shown in the representative blots for male (B) and female mice (C) (left lanes: WT; right lanes: AMPKγ1^H151R). Results are means ± SE. *P < 0.05 and ***P < 0.001 vs. respective WT mice; n = 4–11 mice.

Fig. 4.

Expression and abundance of proteins involved in glucose metabolism. Glucose transporter (GLUT)1 and GLUT4 protein abundance in tibialis anterior muscle from 22-wk-old male (A and B) and female (D and E) WT (open bars) and AMPKγ1^H151R transgenic mice (black bars) are shown in representative Western blots (left lanes: WT; right lanes: AMPKγ1^H151R). Quantification is presented in AU. mRNA expression of Pdk4 was determined in EDL muscle of 22-wk-old WT and AMPKγ1^H151R transgenic male (C) and female (F) mice and is shown as fold change to expression in WT mice. Results are means ± SE. *P < 0.05, **P < 0.01, and ***P < 0.001; n = 4–11 mice.

AMPK activation drives whole body carbohydrate metabolism.

Food intake was increased (13%) in male but not female AMPKγ1^H151R transgenic mice (Fig. 5, A and D). This change in food intake was most pronounced during the active phase (Fig. 5B). Total locomotor activity was reduced in both male (24%) and female (21%) AMPKγ1^H151R transgenic mice on two of three measured axes (X: long side of the cage; Y: short side of the cage; Z: height above the animal; Fig. 5, C and F). These changes were not caused by reductions in movement to the feeder, as the total number of meals was not altered (data not shown). Although food intake was increased and total locomotor activity was reduced in male AMPKγ1^H151R transgenic mice, body weight was unaltered (Table 1). V̇o₂ and V̇co₂ were increased during day and night in male (13–15%) but not female AMPKγ1^H151R transgenic mice (Fig. 6, A, B, E, and F). RER was increased during the day and night in males (10–13%) and females (10%), indicating increased reliance on carbohydrates, rather than fats, as a fuel source (Fig. 6, C and G). RER was reduced to a similar extent in male and female AMPKγ1^H151R and wild-type mice during an overnight fast (male wild types: 0.75 ± 0.01; γ1^H151R: 0.75 ± 0.01; and female wild types: 0.75 ± 0.01; γ1^H151R: 0.76 ± 0.01). Heat production was slightly increased during the daytime in male transgenic mice (2.6%) but significantly decreased in female (9–13%) AMPKγ1^H151R transgenic mice during the day and night (Fig. 6, D and H).

Fig. 5.

Food intake and locomotion. Food intake and locomotion were measured for 2 consecutive days in mice (16–22 wk of age) housed in metabolic cages. Total food intake of male (A) and female (D) WT (open bars) and AMPKγ1^H151R transgenic mice (black bars) and food intake during the light and dark phases of male (B) and female (E) mice during the same periods. Locomotion of male (C) and female (F) WT and transgenic mice was measured by laser beam breaks on the x-axis (X; long side of the cage, black bars), y-axis (Y; short side of the cage, open bars), and z-axis (Z; height above the animal, gray bars). Results are means ± SE. *P < 0.05, ^xP < 0.05 (x-axis), ^yP < 0.05 (y-axis), and ^zP < 0.05 (z-axis) vs. respective WT mice; n = 6–8 mice.

Table 1.

Body weight and tissue mass

	Male		Female
	WT	γ1H151R	WT	γ1H151R
Body weight, g	30.6 ± 1.0	30.9 ± 1.5	22.1 ± 0.5	21.9 ± 0.5
Fat mass, g	3.5 ± 0.5	3.6 ± 0.6	2.5 ± 0.2	2.2 ± 0.2
Lean mass, g	26.0 ± 1.0	25.9 ± 1.2	18.7 ± 0.3	18.4 ± 0.4

Results are presented as means ± SE; n = 7–9 mice. WT, wild type. Body weight, total body fat, and lean mass were measured in 19-wk-old AMP-activated protein kinase (AMPK)γ1^H151R transgenic and WT mice using Echo-MRI.

Fig. 6.

Calorimetric parameters. O₂ consumption (V̇o₂), CO₂ production (V̇co₂), respiratory exchange ratio (RER), and heat were measured for 2 consecutive days in mice (16–22 wk of age) housed in metabolic cages. V̇o₂, V̇co₂, RER, and heat of male (A–D) and female (E–H) WT (open bars) and AMPKγ1^H151R transgenic (black bars) mice are shown during the light and dark phases. Results are means ± SE. *P < 0.05 and ***P < 0.001 vs. respective WT mice; n = 6–8 mice.

Skeletal muscle AMPKγ1^H151R reduces WAT mass in female mice.

Body weight was similar between AMPKγ1^H151R transgenic and wild-type mice at 19 wk of age. Total lean and fat mass were unaltered in male and female AMPKγ1^H151R transgenic mice (Table 1). However, total perigonadal WAT weight was reduced by 38% in female but not male AMPKγ1^H151R transgenic mice (Fig. 7, A and F), indicating sex-specific effects on fat storage. Likewise, serum leptin levels were reduced by 52% in female (Fig. 7G) but not male (Fig. 7B) AMPKγ1^H151R transgenic mice. Such a reduction in serum leptin concentration is classically associated with either increased leptin sensitivity or increased appetite and food intake. Although we did not directly test leptin sensitivity, food intake was unaltered between female AMPKγ1^H151R transgenic and wild-type mice (Fig. 5, D and E). Interestingly, sex-dependent changes in gene expression were observed in perigonadal WAT and BAT from female AMPKγ1^H151R transgenic mice. A significant increase in uncoupling protein-1 (Ucp-1) mRNA in BAT (24%) and a tendency for increased mRNA in perigonadal WAT (5.7-fold, P = 0.087) was observed in female (Fig. 7, H and I) but not male AMPKγ1^H151R transgenic mice (Fig. 7, C and D). Furthermore, β₃-adrenergic receptor mRNA was increased (75%) in perigonadal WAT from female AMPKγ1^H151R transgenic mice (Fig. 7J) but unaltered in male mice (Fig. 7E). Collectively, these findings suggest that skeletal muscle-specific overexpression of AMPKγ1^H151R alters gene expression in both brown and white perigonadal adipose depots. Skeletal muscle triglyceride level was not altered in either male or female AMPKγ1^H151R transgenic vs. wild-type mice (data not shown).

Fig. 7.

White adipose tissue (WAT) weight, leptin levels, and fat tissue gene expression. Gonadal fat pad weight of 4-h-fasted male (A) and female (F) WT (open bars) and AMPKγ1^H151R transgenic mice (black bars) (22 wk of age) was measured after tissue dissection. Serum leptin was measured in the same male (B) and female (G) mice. mRNA expression of uncoupling protein (Ucp1) in gonadal WAT and brown adipose tissue (BAT) of male (C and D) and female (H and I) mice and β₃-adrenergic receptor (Adrb3) in the gonadal WAT of the same mice (E and J) is shown as fold change relative to expression in WT mice. Results are shown as means ± SE. *P < 0.05; n = 6–9 mice.

Increased insulin-mediated glucose utilization upon AMPK activation in female and male mice.

Whole body glucose oxidation was increased in AMPKγ1^H151R transgenic mice (Fig. 6, C and G), suggesting increased peripheral glucose uptake and glucose utilization. Male glucose tolerance and serum insulin levels during the glucose tolerance test were unaltered between AMPKγ1^H151R transgenic and wild-type mice (Fig. 8, A–C). However, insulin levels were reduced in female AMPKγ1^H151R transgenic mice at baseline (57%) and 15 min after the injection of glucose (40%) (Fig. 8F). These results suggest the female AMPKγ1^H151R transgenic mice have increased insulin sensitivity since lower levels of insulin were sufficient to achieve normal glucose tolerance (Fig. 8, D and E). Interestingly, glucose utilization was increased in both male (35%) and female (25%) AMPKγ1^H151R transgenic mice during the hyperinsulinemic euglycemic clamp at an insulin infusion of 1.75 mU·kg⁻¹·min⁻¹ (Fig. 9, A and D). Glucose infusion rates tended to be increased in female AMPKγ1^H151R transgenic mice (P = 0.07; Fig. 9E). The suppression of hepatic glucose production was similar between both male and female AMPKγ1^H151R transgenic and wild-type mice (Fig. 9, C and F). Blood glucose and insulin levels during the hyperinsulinemic euglycemic clamp were similar between both male and female AMPKγ1^H151R transgenic and wild-type mice (Table 2). Collectively, skeletal muscle-specific overexpression of AMPKγ1^H151R enhances peripheral insulin sensitivity and energy expenditure and decreases WAT mass in a sex-specific manner (Table 3).

Fig. 8.

Glucose tolerance and insulin sensitivity. Intraperitoneal glucose tolerance tests (IPGTT) were performed in 4-h-fasted WT (●) and AMPKγ1^H151R transgenic (□) male (A) and female (D) mice (at 17 wk of age). Total area under the curve (AUC) is shown for WT (open bars) and AMPKγ1^H151R transgenic (black bars) male (B) and female (E) mice. Plasma insulin levels were measured before and 15 min after the glucose injection in male (C) and female (F) mice. Results are means ± SE. *P < 0.05 vs. respective WT mice; n = 4–9 mice.

Fig. 9.

Hyperinsulinemic euglycemic clamp. Whole body glucose utilization was assessed in conscious 17-wk-old male (A) and female (D) WT (open bars) and AMPKγ1^H151R transgenic mice (black bars) at basal and clamped states. Glucose infusion rate (GIR) of 30% glucose solution and suppression of hepatic glucose production (HGP) through insulin infusion (1.75 mU·kg⁻¹·min⁻¹) are presented for male (B and C) and female (E and F) mice. Results are shown as means ± SE. *P < 0.05 vs. respective WT mice; n = 3–8 mice.

Table 2.

Glucose and insulin levels during the hyperinsulinemic euglycemic clamp

	Male		Female
	WT	γ1H151R	WT	γ1H151R
Body weight, g	30.0 ± 0.9	29.9 ± 1.7	23.6 ± 0.5	22.6 ± 0.4
Basal plasma glucose, mM	8.2 ± 1.9	11.1 ± 0.3	10.0 ± 0.6	10.5 ± 0.8
Basal plasma insulin, ng/ml	0.9 ± 0.4	0.5 ± 0.1	0.4 ± 0.1	0.3 ± 0.03
Clamp plasma glucose, mM	8.2 ± 1.4	11.1 ± 0.2	9.4 ± 0.6	10.3 ± 0.3
Clamp plasma insulin, ng/ml	1.9 ± 0.4	1.5 ± 0.4	1.1 ± 0.1	1.1 ± 0.1

Results are means ± SE; n = 3–8 mice. Whole body glucose utilization was assessed in AMPKγ1^H151R transgenic and WT mice by a hyperinsulinemic euglycemic clamp.

Table 3.

Summary of sex-specific responses to the AMPKγ1^H151R transgene

	Male	Female
Body weight	↔	↔
Body composition	↔	↔
Lean mass	↔	↔
Fat mass	↔	↔
Glucose tolerance	↔	↔
Glucose utilization	↑ (35%)	↑ (25%)
Insulin sensitivity (clamp)	↑	↑
RER	↑ (10–13%)	↑ (10%)
V̇o2 and V̇co2	↑ (13–15%)	↔
Heat	↑ (3%)	↓ (9–13%)
Food intake	↑ (14%)	↔
Voluntary locomotion	↓ (24%)	↓ (21%)
Glycogen (gastrocnemius)	↑ (13.5-fold)	↑ (10.7-fold)
Glycogen (heart)	↔	↔
Perigonadal WAT mass	↔	↓ (38%)
Serum leptin	↔	↓ (51%)
Ucp1 mRNA in BAT	↔	↑ (24%)
Adrb3 mRNA in WAT	↔	↑ (75%)

RER, respiratory exchange ratio; V̇o₂, O₂ consumption; V̇co₂, CO₂ production; WAT, white adipose tissue; BAT, brown adipose tissue; Ucp1, uncoupling protein 1; Adrb3, β₃-adrenergic receptor. Arrows indicate the direction of change compared with WT mice (↑upregulated; ↓downregulated; ↔unchanged). %Change between AMPKγ1^H151R transgenic vs. respective WT mice is shown in parentheses.

DISCUSSION

Pharmacological or genetic modulation of AMPK in skeletal muscle regulates glucose and lipid metabolism and improves metabolic disturbances associated with obesity and insulin resistance. Previously, we have reported that transgenic mice overexpressing a mutant form (AMPKγ3^R225Q) of the dominantly expressed AMPKγ3 regulatory subunit in skeletal muscle have increased glycogen content, enhanced skeletal muscle energetics, and metabolic flexibility and are protected against the development of dietary-induced insulin resistance (5, 7). Given that the three mammalian γ-subunits are variable in the NH₂-terminal sequence, the degree of activation and the subsequent effects on AMPK signals controlling diverse metabolic responses may depend on the specific isoform present in the AMPK complex (1, 27). Although the γ₃-subunit is the predominant form in skeletal muscle, the more ubiquitously expressed γ₁ regulatory subunit also contributes to AMPK activation and is potentially more sensitive to pharmacological AMPK activators (22). Here, we tested the hypothesis that overexpression of a single mutation in the AMPKγ1 isoform (AMPKγ1^H151R) in skeletal muscle improves whole body glucose metabolism and energy homeostasis. Our study is distinguished from earlier approaches by the fact that we identified sex-dependent peripheral effects of skeletal muscle-specific AMPKγ1 isoform activation.

Chronic AMPK activation in genetically modified mouse models, including skeletal muscle-specific AMPKγ3^R225Q and AMPKγ1^R70Q transgenic mice, leads to glycogen accumulation (6, 8). Consistently, we observed the overexpression of AMPKγ1^H151R causing increased glycogen content in skeletal muscle. The mechanism by which AMPK activation leads to glycogen accumulation under these conditions remains unclear but may involve increased glucose uptake. Acute AMPK stimulation by 5-amino-1-β-d-ribofuranosyl-imidazole-4-carboxamide treatment increases skeletal muscle glucose uptake (29) and glycogen synthase activity via increased levels of glucose 6-phosphate (21). However, in AMPKγ1^H151R transgenic mice, glycogen synthase in skeletal muscle is highly phosphorylated and inhibited under fasted conditions, supporting the notion that decreased glycogenolysis, rather than increased synthesis, promotes glycogen accumulation upon chronic AMPK activation. Based on the available results from AMPKγ3^R225Q, AMPKγ1^R70Q, and AMPKγ1^H151R transgenic mouse models, the increase in skeletal muscle glycogen is likely a consequence of the profound AMPK activation rather than the presence of specific γ-isoforms.

Skeletal muscle energetics are improved by AMPK activation. We have reported previously that isolated skeletal muscle from AMPKγ3^R225Q transgenic mice is fatigue resistant in response to ex vivo contraction (5). Moreover, voluntary wheel-running performance is increased in AMPKγ1^R70Q transgenic mice (35). Given that skeletal muscle glycogen is increased in AMPKγ3^R225Q and AMPKγ1^R70Q transgenic mice, the increased substrate availability for the working muscle may account for the improvement in energetics. Conversely, voluntary locomotor activity is decreased in AMPKγ1^H151R transgenic mice. This result was unexpected given the earlier evidence of improved functional properties and work capacity in AMPKγ3^R225Q and AMPKγ1^R70Q transgenic mice. However, the degree of AMPK activation and increased glycogen content in the AMPKγ1^H151R transgenic mice was far greater than in AMPKγ3^R225Q and AMPKγ1^R70Q transgenic mice, and this level of AMPK activation may have a deleterious effect on generalized skeletal muscle function. Thus, AMPK activation may be beneficial up to a point but deleterious if activities are excessive or if muscle substrate demand is not increased by strenuous exercise.

AMPK activation has been linked to metabolic flexibility and the shift between glucose and lipids as fuel sources under fed and fasting conditions, respectively. Here, we report that skeletal muscle overexpression of AMPKγ1^H151R alters whole body substrate utilization. The moderate increase in respiratory exchange ratio of chow-fed male and female AMPKγ1^H151R transgenic mice indicates a shift toward increased glucose metabolism rather than fatty acid oxidation under fed conditions. Paradoxically, we find increased ACC phosphorylation in isolated skeletal muscle from male and female AMPKγ1^H151R transgenic mice, which is consistent with AMPK activation but suggestive of increased fatty acid oxidation. However, during strenuous exercise, glucose oxidation is increased (37), concomitant with increased AMPK activation (40). Under fasting conditions, AMPKγ1^H151R mice switch to fatty acid oxidation, similarly to wild-type mice, indicating an overall wider range of fuel utilization. We cannot exclude that the slight increase in food intake may contribute to this shift in fuel utilization in male AMPKγ1^H151R transgenic mice. However, food intake was unaltered in female mice despite enhanced glucose oxidation. The increase in carbohydrate oxidation in the AMPKγ1^H151R transgenic mice is further supported by findings of increased glycolytic type IIa/x fibers upon AMPKγ1 activation in AMPKγ1^R70Q mice (35). Thus, fiber-type transformation may contribute to the altered fuel utilization.

Sex-dependent differences in body composition were noted in AMPKγ1^H151R transgenic mice. Despite the increase in food intake in the male AMPKγ1^H151R transgenic mice, body weight was unaltered, which was likely due to the increased energy expenditure represented by increased V̇o₂ and V̇co₂. In female AMPKγ1^H151R transgenic mice, total body weight was also unaltered, but perigonadal white adipose tissue mass was reduced compared with wild-type mice. Consistent with the reduction in fat mass, circulating leptin levels were reduced in female AMPKγ1^H151R transgenic mice. The reduction of serum leptin in the female but not male AMPKγ1^H151R transgenic mice may exert a feedback mechanism to control food intake and adiposity. The mechanisms for these sex-dependent effects of AMPK activation on body composition are unknown but may involve alterations in estrogen or testosterone signaling or differences in leptin sensitivity. Previous studies on transgenic AMPK mouse models have focused mainly on male mice. Thus, it is striking that the phenotypic effects of AMPKγ1 activation observed in this study are dependent partly upon sex, with several whole body changes observed only in female mice. To explore the sex difference in adiposity, we determined mRNA expression of UCP1 in BAT and the β₃-adrenergic receptor in perigonadal WAT. In female AMPKγ1^H151R transgenic mice, mRNA expression of UCP1 in BAT and the β₃-adrenergic receptor in perigonadal WAT is increased, concomitant with reduced white adipose mass. However, V̇o₂ and V̇co₂ are unaltered, whereas heat production is reduced, indicating reduced energy expenditure in female AMPKγ1^H151R transgenic mice. Sex-specific characteristics of skeletal muscle morphology or myokine secretion may account for differences in phenotypic effects of the AMPK activation. Indeed, submaximal exercise was shown to increase free AMP and AMPK activity in skeletal muscle in men, but not in women, which was potentially due to improved cellular energy balance in women (36).

Skeletal muscle-specific overexpression of the mutant AMPKγ1^H151R isoform influences whole body metabolism. Insulin levels in female AMPKγ1^H151R transgenic mice are reduced under fasting conditions during a glucose tolerance test, which is indicative of increased insulin sensitivity. In humans, insulin sensitivity is generally greater in females compared with males (26). Moreover, in obese mouse models, estrogen signaling improves hepatic insulin sensitivity (45). Estrogen increases AMPK and Akt phosphorylation in C₂C₁₂ myotubes (14) as well as 3T3-L1 adipocytes (23), suggesting direct sex hormone effects on AMPK signaling. Importantly, our hyperinsulinemic euglycemic clamp studies reveal that whole body glucose utilization is increased in both female and male AMPKγ1^H151R transgenic mice, supporting a role for AMPK activation to enhance insulin sensitivity and whole body metabolism.

In conclusion, AMPK mutations, including AMPKγ1^R70Q and AMPKγ3^R225Q, located in the CBS1 domain as well as AMPKγ1^H151R in the CBS2 domain lead to increased skeletal muscle glycogen content and additional metabolic changes that are indicative of improved carbohydrate metabolism. Skeletal muscle-specific overexpression of AMPKγ1^H151R results in alterations in whole body energy homeostasis, including increased locomotion, energy expenditure, and insulin sensitivity, as well as sex-specific reductions in perigonadal white adipose tissue mass and serum leptin (in female transgenic mice). Thus, AMPK activation by modifying γ₁- or γ₃-subunits may protect against the development of insulin resistance in obesity or type 2 diabetes by enhancing insulin sensitivity and altering metabolic flexibility.

GRANTS

This study was financed by grants from the Swedish Research Council, the European Research Council Advanced Grant Ideas Program, the Novo Nordisk Foundation, the European Foundation for the Study of Diabetes, the Swedish Foundation for Strategic Research, the Swedish Diabetes Foundation, and the Strategic Diabetes Program at Karolinska Institutet.

DISCLOSURES

The authors have no relevant financial or nonfinancial relationships to disclose.

AUTHOR CONTRIBUTIONS

M.S., M.G.M.J., J.R.Z., and M.B. conception and design of research; M.S. and M.B. performed experiments; M.S. and M.B. analyzed data; M.S., J.R.Z., and M.B. interpreted results of experiments; M.S. prepared figures; M.S. and M.B. drafted manuscript; M.S., M.G.M.J., J.R.Z., and M.B. edited and revised manuscript; M.S., M.G.M.J., J.R.Z., and M.B. approved final version of manuscript.